<ul><li>GitLab’s AI-powered coding assistant, Duo, faced a significant security issue due to prompt injection vulnerability.</li><li>Attackers could manipulate Duo by embedding hidden prompts in code comments, commit messages, or merge-request descriptions.</li><li>Exploiting this vulnerability could lead to harmful code changes, redirection to malicious websites, or leakage of sensitive data.</li><li>GitLab has patched the flaws, emphasizing the importance of maintaining security in AI-integrated development workflows.</li></ul>

GitLab Duo AI Vulnerability Exposes Developers to Code Theft

Discover more